9:00 a.m., panels on the bench
Nine panels, same lot, same model. They are not identical. Two read about 200 K warmer. One is 4 percent dimmer. None of that is a defect; it is normal manufacturing spread, and it is exactly what calibration exists to remove.
9:20, warm-up
The panels run a full white slide for 30 minutes before anyone measures anything. LED backlights and LCD panels both shift colour as they warm up. Measuring a cold panel and calibrating it is how you get a wall that looks right for the first hour and wrong afterwards.
9:50, measurement
A colorimeter is placed at nine points per panel, 81 readings for the wall. The system records white point, luminance, and gamma for each. The target is not the brightest panel; it is a common target slightly below the average, so every panel can be brought to it by reducing output rather than by pushing one beyond its limit.
10:40, correction
Each panel gets a lookup table that adjusts RGB gain. Because the correction only reduces, the whole wall ends up at the brightness of the dimmest panel. If the spread is 6 percent, you lose 6 percent of brightness. That is the trade: uniformity costs a little light.
11:30, the check nobody skips
The wall is displayed as a full white field, then a 20 percent grey field, then a colour ramp, then a real photograph. Three people look at it from four metres. If anyone can pick out a panel in under five seconds, the process goes back a step.
What ships with the wall
A calibration report per panel with the measured values before and after, the target white point, and the date. Ask for it. A supplier who cannot produce this document did not do the work, no matter what the sales sheet says.
Why it matters more than the panel grade
Two walls built from the same B-grade stock, one calibrated and one not, will look dramatically different. Calibration cannot fix a dead pixel, but it can make a mixed batch look like a matched set. It is the cheapest quality upgrade in the whole project.